A5070058501- Multimodal Machine Learning Applications
- Digital Innovation in Industries
- Video Analysis and Summarization
- Domain Adaptation and Few-Shot Learning
- Advanced Image and Video Retrieval Techniques
- Corporate Governance and Management
- Human Pose and Action Recognition
- Digital Platforms and Economics
- Aesthetic Perception and Analysis
- Music and Audio Processing
- Visual Attention and Saliency Detection
- Distributed systems and fault tolerance
- Topic Modeling
- Misinformation and Its Impacts
- Blockchain Technology Applications and Security
- Matrix Theory and Algorithms
- Mobile Agent-Based Network Management
- Business Process Modeling and Analysis
- Music Technology and Sound Studies
- Digital Media and Visual Art
- Educational Technology and Optimization
- Advanced Computational Techniques and Applications
- Education, Innovation and Language Studies
- Media Influence and Health
- Wireless Sensor Networks and IoT
University of International Business and Economics
2022-2025
Abstract How to raise donations effectively, especially in the E‐era, has puzzled fundraisers and scientists across various disciplines. Our research focuses on donation‐based crowdfunding projects investigates how emotional valence expressed verbally (in textual descriptions) visually facial images) project descriptions affects performance. Study 1 uses field data ( N = 3817), grabs information from a top platform, computes visual verbal using deep‐learning‐based affective computing method...
Video moment retrieval aims at localizing a specific from an untrimmed video by sentence query. Most methods rely on heavy annotations of moment-query pairs. Recent zero-shot reduced annotation cost, yet they neglected the global visual feature due to separation and text learning process. To avoid lack features, we propose Prompt-based Zero-shot Moment Retrieval (PZVMR) method. Motivated frame prompt learning, design two modules: 1) Proposal Prompt (PP): We randomly masks sequential frames...
Image aesthetic assessment is an important issue in multimedia, but most existing studies employ supervised learning methods that rely on large-scale annotated data. However, scoring annotations are difficult to obtain large quantities. Therefore, this paper explores zero-shot image assessment. We predict scores by introducing knowledge of different attributes (e.g., Focus). First, we use prompt tuning a unique for each attribute as external knowledge. Second, leverage relations considering...
Video moment retrieval (VMR) is a cutting-edge vision-language task locating segment in video according to the query. Though methods have achieved significant performance, they assume that training and testing samples share same action types, hindering real-world application. In this paper, we specifically consider new problem: by queries with unseen actions. We propose plug-and-play structure, Routing Evidence (RE), multiple evidence-learning heads dynamically route one locate sentence an...
Video moment retrieval locates a specified by sentence query. Recent approaches have made remarkable advancements with large-scale video-sentence annotations. These annotations require extensive human labor and expertise, leading to the need for unsupervised fashion. Generating pseudo-supervision from videos is an effective strategy. With power of pre-trained model, we introduce knowledge into constructing pseudo-supervision. The main technical challenge improving diversity alleviating noise...
In recent years, learnable prompts have emerged as a major prompt learning paradigm, enhancing the performance of large-scale vision-language pre-trained models in few-shot image classification. However, methods are often time-consuming and inflexible because 1) class-specific inefficient certain situations; 2) instance-specific put fixed position. To address these issues, inspired by coarse-to-fine decision-making paradigm human, we propose an Instance-Aware Hierarchical-Structured Policy...